SEBI proposes guidelines for AI usage in securities markets
What's the story
The Securities and Exchange Board of India (SEBI) has proposed a set of guiding principles for the responsible use of Artificial Intelligence (AI) and Machine Learning (ML) in securities markets.
The regulator has also suggested a "regulatory lite" framework for using AI/ML in the securities market for purposes other than business operations directly impacting customers.
Risk management
Proposed guidelines aim to ensure investor protection, market integrity
The proposed guiding principles by SEBI are aimed at maximizing benefits while minimizing potential risks from the use of AI/ML-based applications in securities markets.
The intention is to ensure investor protection, market integrity, and financial stability, SEBI said in its consultation paper
Currently, market participants mainly use AI/ML for advisory services, risk management, client identification and monitoring, surveillance purposes, as well as pattern recognition and internal compliance purposes.
System management
Responsible management of AI/ML systems is crucial: SEBI
SEBI has stressed the importance of responsible management of AI/ML systems, given their potential to create or exacerbate certain risks.
These risks could affect the efficiency of financial markets and have adverse effects on investors.
To address this, SEBI has proposed high-level principles to guide market participants in establishing reasonable procedures and control systems for supervising and governing the use of AI/ML applications or tools.
Monitoring teams
Working group recommends establishment of internal monitoring team
A working group constituted by SEBI has suggested that market participants using AI/ML models should have an internal team with the necessary skills and experience.
This team would be responsible for monitoring the performance, efficacy, and security of algorithms deployed throughout their lifecycle.
They would also ensure auditability and explain the interpretability of such models as part of their proposal to regulate AI/ML in securities markets.
Model usage
Participants should disclose use of AI/ML models to customers
The working group has also recommended that market participants using AI/ML models for business operations should disclose this information to their respective customers.
This is to promote transparency and accountability.
Further, it has been suggested that these participants should adequately test and monitor the AI/ML-based models continuously.
The testing should take place in a segregated environment from the live one before deployment to ensure the expected behavior of these models in stressed and unstressed market conditions.